Github user jkbradley commented on a diff in the pull request: https://github.com/apache/spark/pull/2595#discussion_r18297121 --- Diff: mllib/src/main/scala/org/apache/spark/mllib/tree/DecisionTree.scala --- @@ -518,30 +512,69 @@ object DecisionTree extends Serializable with Logging { agg } - // Calculate bin aggregates. - timer.start("aggregation") - val binAggregates: DTStatsAggregator = { - val initAgg = if (metadata.subsamplingFeatures) { - new DTStatsAggregatorSubsampledFeatures(metadata, treeToNodeToIndexInfo) - } else { - new DTStatsAggregatorFixedFeatures(metadata, numNodes) + /** + * Get node index in group --> features indices map, + * which is a short cut to find feature indices for a node given node index in group + * @param treeToNodeToIndexInfo + * @return + */ + def getNodeToFeatures(treeToNodeToIndexInfo: Map[Int, Map[Int, NodeIndexInfo]]) + :Map[Int, Option[Array[Int]]] = { + val mutableNodeToFeatures = new mutable.HashMap[Int, Option[Array[Int]]]() + treeToNodeToIndexInfo.values.foreach { nodeIdToNodeInfo => + nodeIdToNodeInfo.values.foreach { nodeIndexInfo => + mutableNodeToFeatures(nodeIndexInfo.nodeIndexInGroup) = nodeIndexInfo.featureSubset + } } - input.treeAggregate(initAgg)(binSeqOp, DTStatsAggregator.binCombOp) + mutableNodeToFeatures.toMap } - timer.stop("aggregation") // Calculate best splits for all nodes in the group timer.start("chooseSplits") + // In each partition, iterate all instances and compute aggregate stats for each node, + // yield an (nodeIndex, nodeAggregateStats) pair for each node. + // After a `reduceByKey` operation, + // stats of a node will be shuffled to a particular partition and be combined together, + // then best splits for nodes are found there. + // Finally, only best Splits for nodes are collected to driver to construct decision tree. + val nodeToFeatures = getNodeToFeatures(treeToNodeToIndexInfo) + val nodeToFeaturesBc = input.sparkContext.broadcast(nodeToFeatures) + val nodeToBestSplits = + input.mapPartitions { points => + // Construct a nodeStatsAggregators array to hold node aggregate stats, + // each node will have a nodeStatsAggregator + val numNodes = nodeToFeatures.keys.size --- End diff -- numNodes has already been computed above; reuse that value in this closure?
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org